-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/46407/#review129955
-----------------------------------------------------------




src/examples/balloon_executor.cpp (lines 45 - 46)
<https://reviews.apache.org/r/46407/#comment193512>

    s/The data of the `TaskInfo`/`TaskInfo.data`/



src/examples/balloon_executor.cpp (lines 49 - 50)
<https://reviews.apache.org/r/46407/#comment193513>

    might detect that and destroy the container before the executor can send a 
TASK_FINISHED update.



src/examples/balloon_executor.cpp 
<https://reviews.apache.org/r/46407/#comment193516>

    I don't quite follow what the bug was previously that you are trying to fix 
here with a separate thread?


- Vinod Kone


On April 19, 2016, 9:51 p.m., Joseph Wu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/46407/
> -----------------------------------------------------------
> 
> (Updated April 19, 2016, 9:51 p.m.)
> 
> 
> Review request for mesos, Greg Mann, Artem Harutyunyan, Kevin Klues, and 
> Vinod Kone.
> 
> 
> Bugs: MESOS-5174
>     https://issues.apache.org/jira/browse/MESOS-5174
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Fixes the logic in the example `balloon-executor` to work in cases not
> exercised by the `ROOT_CGROUPS_BalloonFramework` test.
> 
> * The "task" logic was moved into a separate thread.  This fixes the 
>   case where the balloon executor does not exceed the memory limit.
>   (i.e. by Unblocking the driver's thread while sending status updates.)
> * Changed log messages to use glog.
> * Added logic to prevent multiple task launches with the same executor.
> 
> 
> Diffs
> -----
> 
>   src/examples/balloon_executor.cpp 108ebd9afec5b2d592ffbe5c150a9271f1899f2c 
> 
> Diff: https://reviews.apache.org/r/46407/diff/
> 
> 
> Testing
> -------
> 
> See next review in this chain.
> 
> 
> Thanks,
> 
> Joseph Wu
> 
>

Reply via email to